Most cancer drugs are effective in some patients, but not in others. This results from genetic variation among tumors, and can be observed even among tumors within the same patient. Variable patient response is particularly pronounced with respect to targeted therapeutics. Therefore, the full potential of targeted therapies cannot be realized without suitable tests for determining which patients will benefit from which drugs. According to the National Institutes of Health (NIH), the term “biomarker” is defined as “a characteristic that is objectively measured and evaluated as an indicator of normal biologic or pathogenic processes or pharmacological response to a therapeutic intervention.” (Biomarkers Definitions Working Group, 2001, Clin. Pharmacol. Ther. 69:89-95)
The development of improved diagnostics based on the discovery of biomarkers has the potential to accelerate new drug development by identifying, in advance, those patients most likely to show a clinical response to a given drug. This would significantly reduce the size, length and cost of clinical trials. Technologies such as genomics, proteomics and molecular imaging currently enable rapid, sensitive and reliable detection of specific gene mutations, expression levels of particular genes, and other molecular biomarkers. In spite of the availability of various technologies for molecular characterization of tumors, the clinical utilization of cancer biomarkers remains largely unrealized because few cancer biomarkers have been discovered. For example, a recent review article states:                There is a critical need for expedited development of biomarkers and their use to improve diagnosis and treatment of cancer. (Cho, 2007, Molecular Cancer 6:25)Another recent review article on cancer biomarkers contains the following comments:        The challenge is discovering cancer biomarkers. Although there have been clinical successes in targeting molecularly defined subsets of several tumor types—such as chronic myeloid leukemia, gastrointestinal stromal tumor, lung cancer and glioblastoma multiforme—using molecularly targeted agents, the ability to apply such successes in a broader context is severely limited by the lack of an efficient strategy to evaluate targeted agents in patients. The problem mainly lies in the inability to select patients with molecularly defined cancers for clinical trials to evaluate these exciting new drugs. The solution requires biomarkers that reliably identify those patients who are most likely to benefit from a particular agent. (Sawyers, 2008, Nature 452:548-552, at 548)Comments such as the foregoing illustrate the recognition of a need for the discovery of clinically useful biomarkers and diagnostic methods based on such biomarkers.        
There are three distinct types of cancer biomarkers: (1) prognostic biomarkers, (2) predictive biomarkers, and (3) pharmacodynamic (PD) biomarkers. A prognostic biomarker is used to classify a cancer, e.g., a solid tumor, according to aggressiveness, i.e., rate of growth and/or metastasis, and refractiveness to treatment. This is sometimes called distinguishing “good outcome” tumors from “poor outcome” tumors. A predictive biomarker is used to assess the probability that a particular patient will benefit from treatment with a particular drug. For example, patients with breast cancer in which the ERBB2 (HER2) gene is amplified are likely to benefit from treatment with trastuzumab (HERCEPTIN®), whereas patients without ERBB2 gene amplification are unlikely to benefit from treatment with trastuzumab. A PD biomarker is an indication of the effect(s) of a drug on its molecular target while the patient is taking the drug. Accordingly, PD biomarkers often are used to guide dosage level and dosing frequency, during the early stages of clinical development of a new drug. For a discussion of cancer biomarkers, see, e.g., Sawyers, 2008, Nature 452:548-552.
Tumors driven by EGFR or HER2 often respond to treatment with inhibitors of EGFR or HER2, but these tumors invariably develop resistance to these inhibitors. At least one mechanism of acquired resistance to anti-EGFR or anti-HER2 treatment is activation of ERBB3 (also known as HER3) signaling. See, e.g., Engelman et al., 2006, Clin. Cancer Res. 12:4372; Ritter et al., 2007, Clin. Cancer Res. 13:4909; Sergina et al., 2007, Nature 445:437. NRG1-induced activation of HER2-ERBB3 heterodimers also has been associated with resistance to EGFR inhibitors (Zhou et al., 2006, Cancer Cell 10:39). Thus, ERBB3 plays an important role in development of drug resistance, as well as being involved in tumor initiation and maintenance, through its heterodimerization with EGFR and HER2. Consequently, there has been interest in development of ERBB3 inhibitors, especially anti-ERBB3 antibodies, since ERBB3 lacks kinase activity.
As with other types of targeted therapy, some, but not all, tumors respond to anti-ERBB3 therapy. Therefore, there is a need for diagnostic methods based on predictive biomarkers that can be used to identify patients with tumors that are likely (or unlikely) to respond to treatment with an ERBB3 inhibitor such as an anti-ERBB3 antibody.